In the past 2 years, the development of video-assisted thoracic surgery (VATS) has allowed thoracoscopy to rapidly evolve from a primarily diagnostic tool to a therapeutic tool with potentially useful applications. As with laparoscopy, the initial enthusiasm for VATS must be tempered by the reality of technical inadequacies, complications, and results that have yet to withstand careful scientific scrutiny. The purpose of this study was to determine the technical feasibility of a VATS lobectomy. Forty-four patients with primary bronchogenic carcinoma were evaluated and accepted as potential candidates for VATS lobectomy. After complete preoperative staging, these patients were in clinical stage I. All patients had normal arterial blood gases and adequate pulmonary function (forced expiratory volume in 1 second > 1.5 L) to tolerate a lobectomy. At the time of operation, 3 patients were found to have N2 disease and were excluded from the study. In 35 of the remaining 41 patients, a successful VATS lobectomy was accomplished through two thoracoscopy ports and a non-rib-spreading 6- to 8-cm ''access'' thoracotomy. There were no major intraoperative complications that necessitated conversion to an open thoracotomy. Mean operative time was 153 +/- 26 minutes. All 35 patients recovered uneventfully with a mean hospital stay of 5.7 +/- 1.6 days. A VATS lobectomy is technically feasible using our approach and is potentially safe. However, major advances in thoracoscopic imaging and instrumentation are necessary before this procedure will have the potential for widespread acceptance. Also, the advantages of VATS lobectomy over accepted surgical techniques will have to be carefully documented in randomized trials.